Port-macppc archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Problems with extra packages on -current?



Finally getting back to poking at macppc stuff.  Currently running
6.99.17 from 20130215 or thereabouts, but there's things I've noticed
since about 6.99.11.

First, compiling large packages, namely lang/perl5, fails with the
following message:

[...]
Processing SpecialCasing.txt
Processing LineBreak.txt
Processing EastAsianWidth.txt
panic: memory wrap at lib/unicore/mktables line 9980, <$file_handle> line 7591.
*** Error code 25

Stop.
make: stopped in /var/tmp/pkgsrc/lang/perl5/work/perl-5.16.2
*** Error code 1
[...]

The "panic" message must be from perl's build system as the system stays
up.

'ulimit -a' shows:

1040 [sysop@bondi:~]$ ulimit -a
time(cpu-seconds)    unlimited
file(blocks)         unlimited
coredump(blocks)     unlimited
data(kbytes)         262144
stack(kbytes)        2048
lockedmem(kbytes)    82488
memory(kbytes)       247464
nofiles(descriptors) 128
processes            160
threads              160
vmemory(kbytes)      unlimited
sbsize(bytes)        unlimited

Although this is on my 256MB iMac, I've had the same problem on my 1GB
G4 Powerbook.  I forget what I did to work around the problem now, but
I'm pretty sure unlimiting one of the above items was involved.


Second, a number of extra packages, once built, crash with various forms
of "memory fault" upon being run.  In particular, 'sudo' produces the
the following message on the console:

trap: pid 21886.1 (sudo): user read DSI trap @ 0xfdff6ffc by 0xfdf6a6ec (DSISR 
0x40000000, err=14)

but doesn't leave a core dump.  The program executed with elevated privilege
runs correctly, but when sudo crashes, it causes anything looking for the
exit status to detect failure.

I'll try to get a ktrace of sudo.  I seem to recall that it incurs SIGSEGV
on a close() call.

Likewise, 'x11/rxvt-unicode' ('urxvt') produces the following on the
console:

trap: pid 8681.1 (urxvt): user read DSI trap @ fdff54fc by 0xfdb24fc0 (DSISR 
0x40000000, err=14)

The backtrace is as follows:

[...]
Core was generated by `urxvt'.
Program terminated with signal 11, Segmentation fault.
#0  0xfdb24fc0 in close () from /usr/lib/libpthread.so.1
(gdb) bt
#0  0xfdb24fc0 in close () from /usr/lib/libpthread.so.1
#1  0x0183816c in ptytty::close_tty() ()
#2  0x0181e80c in rxvt_term::run_command(char const* const*) ()
#3  0x0181ec18 in rxvt_term::init2(int, char const* const*) ()
#4  0x01811244 in main ()
(gdb)

I'll try to rebuild with debugging symbols for more information.

I read the thread about locking and my kernels are compiled with
"options FULL".

Any thoughts or clues?

Thanks.

--
|/"\ John D. Baker, KN5UKS               NetBSD     Darwin/MacOS X
|\ / jdbaker[snail]mylinuxisp[flyspeck]com    OpenBSD            FreeBSD
| X  No HTML/proprietary data in email.   BSD just sits there and works!
|/ \ GPGkeyID:  D703 4A7E 479F 63F8 D3F4  BD99 9572 8F23 E4AD 1645



Home | Main Index | Thread Index | Old Index